The revised application proposes a five-year extension of the Early Steps Multisite Project (ES-M) focused on Understanding and Preventing Childhood Drug Use Risk. The ES-M project is a collaboration of the Child and Family Center at the University of Oregon (Dishion), the University of Pittsburgh (Shaw), and the University of Virginia (Wilson). During its first five years the project exceeded expected recruitment rates (731), maintained high retention rates (85%), successfully engaged families in our family-intervention model, and produced reductions in problem behavior and emotion distress in children from ages 2 to 4 years. We have found that improvements in children's adjustment were mediated by improved family management practices. In the next five years we propose to adapt the family intervention model to provide support for each child's transition into the public school system, and conduct developmental analyses to better understand early emergence of school competence, early deviant peer involvement problem behavior, and emotion distress. We will also address the role of siblings in children's development and as an intervention outcome variable. Biannual intensive assessments are proposed that include home visits and direct observations, as well as yearly assessments of the child's problem behavior and self-regulation in the family context. This assessment builds on an existing multiagent-multimethod assessment strategy across all three sites that has included yearly direct observations as well as measurement of child and family characteristics. We propose that the study will test the effectiveness of a feasible, cost-effective, family-centered intervention that can be initiated within WIC or similar service settings and be continued and maintained in the public school environment. In addition, the data set affords a rich opportunity to link caregiver engagement in the intervention with child and family outcomes. The net effect of this family-centered, early intervention strategy is significant reduction of risk for early-onset problem behavior, improved school competence, reduction in children's mental health disorders, and reductions in risk for drug use from age 2 to 10. Advancements in multivariate data analysis such as general growth mixture modeling will be used to study the environmental factors associated with the formation of health and of high-risk developmental trajectories, as well as to determine the effectiveness of the family intervention in altering those trajectories.